Re: design example



Mauri wrote:
> Hi all,
> i need your help for understanding this design example
>
> Problem: A full bridge speed controller circuit is designed to control
> a 12v motor. The switching frequency must be above the audible limit
> (20kHz). The motor has a total resistance of 0.12 Ohms. Choose
> suitable MOSFETs for the bridge circuit, within a reasonable price
> limit, and suggest any heatsinking that may be required. The ambient
> temperature is assumed to be 25ºC.
>
> Solution: Lets have a look at the IRF3205 and see if it is suitable.
> First the drain current requirement. At stall, the motor will take 12v
> / 0.12 Ohms = 100 Amps. We will first make a guess at the junction
> temperature, at 125ºC We must find what the maximum drain current is
> at 125ºC first. The graph of figure 9 shows us that at 125ºC, the
> maximum drain current is about 65 Amps. Therefore 2 IRF3205s in
> parallel should be capable in this respect.
> How much power will the two parallel MOSFETs be dissipating? Lets
> start with the power dissipation whilst ON and the motor stalled, or
> just starting. That is the current squared times the on-resistance.
> What is RDS(on) at 125ºC? Figure 4 shows how it is derated from its
> front-page value of 0.008 Ohms, by a factor of about 1.6. Therefore,
> we assume RDS(on) will be 0.008 x 1.6 = 0.0128. Therefore PD = 50 x 50
> x 0.0128 = 32 Watts. How much of the time will the motor be either
> stalled or starting? This is impossible to say, so we will have to
> guess. 20% of the time is quite a conservative figure - it is likely
> to be a lot less. Since the power causes heat, and the heat conduction
> is quite a slow process, the effect of power dissipation tends to get
> averaged out over quite long time periods, in the region of seconds.
> Therefore we can derate the power requirement with the quoted 20%, to
> arrive at an average power dissipation of 32W x 20% = 6.4W.
> Now we must add the power dissipated due to switching. This will occur
> during the rise and fall times, which are quoted in the Electrical
> Characteristics table as 100ns and 70ns respectively. Assuming the
> MOSFET driver can supply enough current to fulfill the requirements of
> these figures (gate drive source resistance of 2.5 Ohms = pulse output
> drive current of 12v / 2.5 Ohms = 4.8 Amps), then the ratio of
> switching time to steady-state time is 170ns * 20kHz = 3.4mW which is

I take it you mean micro-seconds there.

> negligable. These on-off timings are a bit crude however, for more
> information about on-off times, see here.
> Now what are the switching requirements? The MOSFET driver ship we use
> will cope with most of these, but its worth checking. The turn-on
> voltage, Vgs(th), from the graphs of Figure 3 is just over 5 Volts. We
> have already seen that the driver should be able to source 4.8 Amps
> for a very short period of time.
> Now what about the heatsink. You may want to read the chapter on
> heatsinks before this section. We want to keep the temperature for the
> semiconductor junction below 125ºC, and we have been told that the
> ambient temperature is 25ºC. Therefore, with a MOSFET dissipating 6.4W
> on average, the total thermal resistance must be less than (125 - 25)
> / 6.4 = 15.6 ºC/W. The thermal resistance from junction to case makes
> up for 0.75 ºC/W of this, typical case to heatsink values (using
> thermal compound) are 0.2 ºC/W, which leaves 15.6 - 0.75 - 0.2 = 14.7
> ºC/W for the heatsink itself. Heatsinks of this θjc value are
> quite small and cheap. Note that the same heatsink can be used for
> both MOSFETs to the left of or to the right of the load in the H-
> bridge, since these two MOSFETs are never both on at the same time,
> and so can never both be dissipating power at the same time. The cases
> of them must be electrically isolated however. See the heatsinks page
> for more information on the required electrical isolation.
>
> These are my doubts:
>
> 1) "Assuming the MOSFET driver can supply enough current to fulfill
> the requirements of these figures (gate drive source resistance of 2.5
> Ohms = pulse output drive current of 12v / 2.5 Ohms = 4.8 Amps)". What
> does it means ? Shouldn't the choise of the mosfet driver be based on
> Qg= I*T --> I=Qg/T --> I=146ns*20KHz = 2,92mA ?

The gate has to be charged and discharged very quickly, by short pulses
taking of the order of 100 nanoseconds or less, each. I think you are using
the wrong value for time in this calculation, but I'm a bit confused by your
use of 'ns' for units of charge. I take it this is meant to be 146
nano-Coulombs total gate charge?



.



Relevant Pages

  • design example
    ... The motor has a total resistance of 0.12 Ohms. ... How much power will the two parallel MOSFETs be dissipating? ... the effect of power dissipation tends to get ... The MOSFET driver ship we use ...
    (sci.electronics.design)
  • Re: MOSFET packaging
    ... The basic idea of switching control is that the MOSFET is either turned on ... Power = current * voltage. ... As for Rdsratings, it seems this follows the same principle as ... lower voltage drop and power dissipation? ...
    (sci.electronics.basics)
  • Re: How to quickly turn on/off N-FET switching high side of 55V?
    ... MOSFET dies sizes by comparing power dissipation ratings ... Wouldn't power dissipation ratings also be highly dependent on the ... package that the FET was in? ...
    (sci.electronics.design)
  • Re: fet rise/fall times
    ... charge and the gate drive currents, but it seems that maybe some ... many OF TODAY'S high-voltage power MOSFETs ... The gate-spreading resistance info you ... mosfet will do is not available. ...
    (sci.electronics.design)
  • Re: H-Bridge blew up 10 MOSFET a day
    ... mosfet chip). ... At 10 amps you will have 4 times the power. ... There is no capacitor directly across the TEC. ... shows a slightly different circuit with 2.2 uF across the TEC. ...
    (sci.electronics.design)